Introduction¶
Raw data must be cleaned and transformed prior to analysis, which typically includes reformatting data, making data modifications, and combining data sets to enhance data. I utilized the tweet archive of Twitter user @dog rates, commonly known as WeRateDogs, for my dataset. WeRateDogs is a Twitter account that scores people's pets and includes a funny comment about the dog. The WeRateDogs Twitter archive includes basic tweet data for all of their 5000+ tweets, but not everything. The archive did not include the retweet and favorite counts for each tweet, so I had to obtain this additional data from the twitter account using the archive's tweet ID.
I also obtained an image predictions file from Udacity servers that had the top three predictions for dog breeds based on images from Twitter. After cleaning the data, I performed some data analysis and visualizations, the results of which are presented below.
Correlation between Retweets and Favorites¶
There is a strong relationship between a tweet's retweet counts and favorite counts, with an rsquared of 0.92; in other words, if a tweet has a high fave count, it will also have a high retweet count. When these two values were plotted against each other, the plot below was obtained, which clearly reveals a high correlation by the highly tight grouping of dots along the angled, straight line.
All tweets have a higher number of favorites than retweets. Tweets with a high favorite count also have a high retweet count and are rated better.
High Ratings¶
The distribution demonstrates how many tweets have really high ratings; in fact, 75 percent of all tweets had incredibly high ratings! falling between 10 and 14, with 12 being the highest.
The most rated dog breed¶
The black and tan coonhound had the highest mean rating of 14.0, suggesting that it was highly rated among the breeds, followed by the giant schnauzer, Bouvier des Flandres, Saluki, and other breeds.
Top 10 dog breeds by tweet count and favorite count¶
According to the bar chart above, the Golden Retriever is the most popular dog, with around 125 tweets. Labrador Retrievers, Pembroke Spaniels, and Chihuahuas come in second, third, and fourth place, respectively. Also, the top 10 dog breeds in terms of favorite counts and retweet counts are the same. People adore retrievers!
Because people adore golden retrievers! I'd like to dig more into it and see how it evolves and changes through time.¶
The amount of golden retriever tweets varied by month, with July having the most and September having the fewest. The most tweets for golden retrievers were in 2016, and they were generally high in volume on Mondays and Saturdays, so if any paid advertising was to be done, Mondays and Saturdays would be a good time.
Where were all these tweets sents from¶
We can observe that the majority of tweets on golden retrievers are sent from an iPhone.
__